TYDRICHOVA Magdalena

Tervetuloa! Ahoj! Welcome!

Who am I?

My name is Magdaléna Tydrichová Abadie and since Octobre 2023, I am a tenure track assistant professor in Computer science at CentraleSupélec, Paris Saclay University. More precisely, I work at MICS Lab , LogiMICS research group. For even more precision, please refer to the section "Where to find me" below.

I am working in the field called Computational Social Choice. My research mostly tackles problems around structured preferences: I have a set of voters who express their preferences on a set of candidates. Given the set of these preferences, I try to identify and understand its underlying structure. Some typical questions I am interested in could be, for instance: Is there any structure behind these preferences? How can we recognize it? What about its interpretation? What information does it provide about voters/candidates? How it can be useful, and used, in practice?

Independently of my current research topics, I am (and I have always been, as I can remember) passionated by fields like combinatorics (and especially Ramsey theory), finite geometry, graph theory (mostly geometric graph theory) or combinatorial game theory. I am also fascinated by topology or group theory.

Where I come from ?

From Octobre 2019 to March 2023, I was a PhD student in Computer science at LIP6, Sorbonne Université in Paris, under the supervision of Olivier SPANJAARD and Bruno ESCOFFIER.

Prior to that, in 2019, I got my Master degree in Computer science and another one in Mathematics. Although my current research focuses entirely on on Computer science topics, I do not forget my mathematical backgroud, stay pationated by several fields of (discrete) mathematics and would be very happy to discuss, or even better collaborate, with you on more mathematical topics (see below for more details)!

Finally, if you want a more fundamental answer on the question "Where I come from", you have probably could noticed that the cover photo had nothing to do with my research. Highly inspired by the poem Hora Říp by Jaroslav Seifert, it shows the place I come from.

As I already said, I am working in the field called Computational Social Choice. Computational Social Choice tackles, from a computational point of view, the problems arising from collective decisions.

My current research revolves around voting theory, and more precisely around structured preferences. Given a set of voters that express their preferences on a set of candidates, I try to understand the underlying structure of this voting space.

In my PhD thesis, I proposed several relaxations of so-called single-peaked preferences, and I studied several variants of so-called multidimensional Euclidean preferences. I mostly focused on problems linked to the recognition of such structures, and on the axiomatic or geometrical properties guaranteed by these structures.

At the moment, I am trying to interpret these structure. Take for instance the single-peakedness. Although we have quiet a good interpretation of single-peaked axis, the way how to interpret the relations between candidates single-peaked on a tree, cycle or even an arbitrary graph becomes much less clear. However, having more insights into the structure could be helpful to, among others, groupe decision making, preference learning or decision-support tools.

... So what does it mean if two candidates are placed "near" one next to other in the structure? Can we deduce some relations on candidates (e.g., dominance relations, similarities, etc.) from a given structure? And if so, we deal here with another issue: sometimes, several different structures explain the same preference profile. Obviously, each of these structures provides a different structural information. The challenging question is hence to identify a structure that "fits the best" our profile.

That is the main question I am trying to answer currently. If you are interested in this king of topics, I will be very very happy to exchange, or better, collaborate with you. Just let me know!